Board Level Placement

Because of complexity in modern printed wiring board (PWB) designs which can consist of a large number of components, the placement of the components has become an extremely time consuming task and may involve tradeoffs of equally important weight. To help a design engineer, placement automation techniques have been developed. Most of them are developed for improving wireability [Che84, Got78, Han72, Oda87, Qui79, Sia87] by either minimizing of the total wire lengths, or minimizing the maximum crossings of a horizontal or vertical cut of the placement. More sophisticated program uses a combination of both conditions [Wip82]. Recently, some CAD programs with a minimized system failure rate were also developed. Particularly, Pecht and Naft [Pec87] presented a placement method for reliability based on random sampling of placement configurations. Dancer and Pecht [Dan89] presented several optimization schemes for reliability and compared their accuracy and computational speed. Mayer [May78] discussed the optimization for reliability based on the thermal design constraints of electronic systems in avionics. Osterman [Ost92] discussed component placement for reliability based on a failure model which incorporated component temperature, base operating temperature, threshold temperature, and changes in temperature. Eliasi et al. [Eli90] presented a Monte Carlo thermal optimization method for component placement on a PCB.